|[Home] [Help] [Databases] [WorldLII] [Feedback]|
Macquarie Law Journal
SEPARATION OF POWERS, ‘TRADITIONAL’ ADMINISTRATION AND RESPONSIVE REGULATION
GEOFF AIRO-FARULLA AND STEVEN WHITE[∗]
In Plaintiff S157/2002 v Commonwealth, the High Court entrenched judicial review of administrative decision-making for jurisdictional error under s 75(v) and possibly s 75(iii) of the Australian Constitution as a ‘fundamental constitutional principle’. These provisions apply to Commonwealth officers and statutory agencies. However, administrative technologies are rapidly changing. Governance techniques increasingly emphasise the role of non-government agencies in making and implementing public policy. These developments do not fit easily into a constitutional landscape that implicitly assumes that public power will be exercised by public officials and statutory agencies. It is therefore significant that the decision in Plaintiff S157 was also based on the separation of powers doctrine, because that doctrine’s reach is not similarly limited by the Constitution’s text to ‘traditional’ administrative agencies.
Nevertheless, as we discuss in Part II, the High Court developed its separation of powers doctrine in the context of a relatively narrow range of administrative agencies, most importantly the industrial conciliation and arbitration system. Only recently has it come to be applied to a broader range of agencies. In Part III, we outline the doctrine’s main features, and evaluate how well it has applied across this broader range of agencies, all still ‘traditional’ administrative agencies, in the sense that they were established and empowered by statute.
However, Commonwealth government policy increasingly reflects a theory of ‘responsive regulation’ emphasising the role of industry self-regulation based, wholly or in part, on non-statutory industry codes. In Part IV, we examine two cases, Attorney-General (Cth) v Breckler and Australian Communications Authority v Viper Communications, in which such schemes have thrown up separation of powers issues, and evaluate how well the High Court’s separation of powers jurisprudence applies to them.
Our evaluative premises are that there is a public interest in effective government, and that the separation of powers should not unduly prevent Parliament from experimenting with its institutional designs and allocation of functions in order to achieve this. At the same time, accountability is an essential element of good institutional design, necessary to ensure that the right decisions are made, as well as that wrong decisions are not made. We also consider that judicial independence is the separation of powers doctrine’s best justification, as it provides an accountability mechanism politically independent from government. Thus, the separation of powers doctrine ought to ensure effective accountability through judicial oversight, whilst maintaining maximum flexibility in the allocation of primary decision-making authority.
The separation of powers doctrine applies most neatly when ministerial departments implement legislation made by Parliament, and are subject to judicial review by the courts. However, when an interventionist state experiments with different types of non-ministerial agencies, or with the functions it allocates to either administrative agencies or courts, it invites constitutional challenges. Those affected by government action are more likely to challenge unconventional institutional designs on the basis that they transgress the separation of powers doctrine’s limits.
During the second half of the nineteenth century, colonial governments had engaged in nation-building activities to such an extent that contemporaries characterised them as ‘colonial socialism’. As in England, developing theories of responsible government put ministerial departments at the administrative system’s heart, but did not preclude experiments with non-ministerial organisations. Indeed, the Australian colonies continued to use other forms of administrative agencies to a much greater degree than in Britain. The Commonwealth, in contrast, was comparatively slow in developing an administrative system outside the core ministerial departments, except in relation to employment regulation, where it was both active and innovative. The High Court’s restrictive interpretation of the Commonwealth’s legislative powers, particularly the corporations power, certainly did not help. But even where the Commonwealth did have legislative power, with respect to interstate commerce, for example, or (later) broadcasting, it seems to have been remarkably reluctant to use non-ministerial agencies to regulate these sectors, compared to overseas developments.
The Commonwealth was also slow to develop the kind of early ‘welfare state’ institutions developing in other liberal democracies. In Australia, the focus was on using centralised wage fixing to maintain incomes, rather than state-provided income support. The social security system was deliberately restricted to a safety net for those not covered by the ‘living wage’ determined by the employment regulation system. Workers’ wages were also protected by restricted immigration, in part intended to limit competition for jobs. Industry and agriculture, in turn, both benefited from high tariff protection. The link between wage protection and tariff protection was explicit in early Commonwealth legislation.
This ‘protection all round’ approach continued after World War One. During the Great Depression, the wage fixing system delivered real wage cuts without breaking social harmony, leading to greater employer acceptance of the system. Castles has argued that ‘protection all round’ was based on a ‘historic compromise’ between organised labour and capital. This argument has been criticised as
a post hoc rationalisation which owes more to social theory than to the actual cut and thrust of historical conflicts over resources ... sectional interests were engaged in a contingent and highly politicised conflict pursuing their own self-interests over ad hoc protectionism.
Certainly, both organised labour and capital were highly opportunistic, frequently resorting to constitutional challenges when it suited them, including on separation of powers grounds, as in Waterside Workers’ Federation v J W Alexander Ltd and Victorian Stevedoring and General Contracting Co Pty Ltd v Dignan.
The other important separation of powers cases in this period were British Imperial Oil Company Ltd v Federal Commissioner of Taxation and Federal Commissioner of Taxation v Munro. These were prompted by the Commonwealth’s first attempts at establishing administrative review mechanisms, in these cases for taxation decisions. Further developments in administrative review did not occur until the 1970s.
World War Two saw the Commonwealth Labor government, emboldened by an expanded defence power, become both much more interventionist and more innovative in its allocation of functions and powers through a series of National Security Regulations. These were subjected to a series of constitutional challenges by regulated entities, again using separation of powers arguments when they were plausible. Post-war, Labor expanded the social security system, but the High Court stymied its more far-reaching democratic socialist agenda using heads of power and s 92 limitations rather than separation of powers arguments. The unadventurous Menzies government did little to prompt constitutional challenges of any sort to its legislation, including on separation of powers grounds. However, the employment regulation system continued to throw up challenges, resulting in the landmark decision of R v Kirby; Ex parte Boilermakers’ Society of Australia in 1955.
Thus, by the 1970s, the High Court had laid down the foundations of its separation of powers jurisprudence, but in a relatively narrow range of contexts. From that point on, the Commonwealth became much more interventionist and innovative, producing more frequent separation of powers challenges, across a much broader spectrum of institutional designs.
The separation of powers is one of the four key constitutional principles implied in the Constitution’s ‘text and structure’, along with federalism, representative government and responsible government. While the Constitution does not explicitly provide that the Commonwealth Parliament cannot confer judicial powers on administrative agencies, or non-judicial powers on Chapter III courts, this has been held to be the necessary consequence of the Constitution’s structural distinction between the legislative, executive and judicial powers of the Commonwealth in Chapters I, II and III.
However, the Constitution does not define ‘judicial power’, and only gives a weak suggestion of what it might include in its specification of different kinds of ‘matters’ in Ch III. It has thus been left to the High Court to determine what judicial power is, which immediately raises a problem of circularity: it is an exercise of judicial power to determine what is ‘judicial power’. Ultimately, the essence of judicial power under the Australian Constitution is no more and no less than the power to say what else it is and is not. Of course, that principle alone provides no guide to the separation of powers’ application in specific cases. The Court has thus identified a series of characteristics to provide a more ‘reasoned’ basis for distinguishing between judicial and non-judicial power. As we show below, in practice, that jurisprudence has been guided more by pragmatism than by any commitment to theoretical purity or conceptual clarity, in particular, a pragmatic commitment to flexibility in institutional design, resulting in a theory that is more about the ‘overlap’ than the ‘separation’ of powers.
1 The Need for Flexibility
A rigid approach to the separation of powers can stifle institutional innovation, forcing institutional design into historically familiar and accepted forms. Indeed, the United States Supreme Court for a time applied its separation of powers doctrine to frustrate the establishment of a range of independent regulatory agencies. It took President Franklin Roosevelt’s threat to stack the court to bring an end to its obstructionist tactics. The High Court referred to this episode in Boilermakers, when the majority commented:
It is not necessary to trace the course of constitutional development in the United States with respect to the separation of powers. It is enough to say that an unfortunate rigidity in the conception of the boundaries between the three great functions of government led for a time to difficulties both of practice and of theory. 
In Australia, by contrast, references in the cases to the need for flexibility and for effective government have been common. In Dignan, which held that the separation of powers did not prevent Parliament from delegating legislative power to administrative agencies, Evatt J said ‘[u]nless the legislative power of the Parliament extends [to the power to delegate], effective government would be impossible’. Dixon J said:
it is one thing to adopt and enunciate a basic rule involving a classification and distribution of powers ..., and it is another to face and overcome the logical difficulties of defining the power of each organ of government, and the practical and political consequences of an inflexible application of their delimitation.
The main door to institutional experimentation has been the theory that there is a ‘borderland in which judicial and administrative functions overlap’. Closely allied is the principle that a ‘function can take its character from the functionary’, to paraphrase Isaacs J in Munro. Thus, as Deane, Dawson, Gaudron and McHugh JJ said in Brandy v Human Rights and Equal Opportunities Commission, ‘there are functions which, when performed by a court, constitute the exercise of judicial power but, when performed by some other body, do not’. The overlap theory does not limit Parliament to an ‘either/or’ choice. It can give a function to both administrators and Ch III courts, to be performed simultaneously, or successively. This approach has been quite successful in delivering the High Court’s stated aims of flexibility and effectiveness but at the expense of conceptual rigour. The question of which kind of body a power can be given to, largely collapses into the question of which kind of body the power has been given to.
The overlap theory was apparent from the very start. In Huddart Parker, the first High Court separation of powers case, the question was whether a statutory power given to the Comptroller-General of Customs to require a person to answer questions and produce documents was an exercise of judicial power. It was argued that this was analogous to the powers of interrogation of judicial bodies, and was therefore a judicial power. Griffith CJ held ‘[m]any such interrogations are no doubt so entrusted [to judicial bodies], but many others, relating to matters of administration, are entrusted to other authorities’.
However, the overlap theory was really established in the next separation of powers case, New South Wales v Commonwealth, through its treatment of powers to adjudicate. In this context, adjudication does not necessarily require either an oral hearing, or an adversarial process. It simply means deciding issues between contending parties who are separate from the decision-making body itself. In Huddart Parker, Griffith CJ had defined judicial power as
the power which every sovereign authority must of necessity have to decide controversies between its subjects, or between itself and its subjects, whether the rights relate to life, liberty or property. The exercise of this power does not begin until some tribunal which has power to give a binding and authoritative decision (whether subject to appeal or not) is called upon to take action.
Taken literally, this broad definition would reserve a wide range of adjudicatory functions to the courts, to the exclusion of administrative agencies. However, in the Wheat Case, the High Court held that ‘adjudicating’ was not the exclusive domain of Ch III courts. The Commonwealth argued that s 101, which provides that Parliament could give the Commission powers of both ‘adjudication and administration’, indicated that it was to have judicial powers. However, the Court held that ‘the function of adjudication is [not] confined to Courts’. Isaacs J said:
though an ‘adjudication’ in the true sense − and as effective and binding as if made by a Court of Justice − may be made by an administrative body, it does not become the adjudication of a Court of Justice. The nature of the power conferred does not alter the character of the body exercising it, and convert an executive body into a strictly judicial body.
Thus, we see the genesis of the principle that ‘[t]he character of the function often takes its colour primarily from the character of the functionary’, as Isaacs J later put it in Munro. The Privy Council endorsed Isaac J’s Munro judgment on appeal, and the overlap theory has been a staple of the High Court’s separation of powers jurisprudence ever since.
3 The Need for Adjudication?
Before turning to look at the overlap theory’s limits, it is important to consider whether adjudication is a necessary feature of judicial power. It has frequently been assumed that this is inherent in Griffith CJ’s reference to deciding controversies between parties. For example, in R v Trade Practices Tribunal, Windeyer J said that the question whether a person is exercising judicial power ‘can ... only arise when a power of adjudication is exercisable by a Commonwealth official or some person appointed for that purpose under Commonwealth law’.
Although this view is intuitively attractive, we believe it is incorrect. It raises a fundamental issue when applied to administrative decision-making. Most primary administrative decision-making does not involve adjudicating controversies between parties. For example, the Immigration Minister (or delegate) does not adjudicate when deciding whether a protection visa applicant satisfies the statutory criteria. If adjudication is an essential element of judicial power, then by definition such primary decision-makers could never impermissibly have been given judicial power.
On the other hand, it does make sense to speak of merits review tribunals such as the Refugee Review Tribunal as ‘adjudicating’ between primary decision-makers and those affected by their decisions. Therefore, such tribunals have at least the potential to have impermissibly been given judicial powers. This is illogical, given those tribunals ‘stand in the shoes’ of the primary decision-maker. The ‘function takes its nature from the functionary’ principle can’t explain this change, as administrative review tribunals are not Ch III courts. It would also be an invitation to abuse, as it would allow the Commonwealth to make primary decision-makers ‘judges in their own cause’, able to answer separation of powers challenges by simply arguing that they do not adjudicate.
In practice, the view that adjudication is an essential element of judicial power has now been foreclosed by the joint judgment in Plaintiff S157, which held that it is a ‘fundamental constitutional principle’ that:
A privative clause cannot operate so as to allow a non-judicial tribunal or other non-judicial decision-making authority to exercise the judicial power of the Commonwealth. Thus, it cannot confer on a non-judicial body the power to determine conclusively the limits of its own jurisdiction.
In context, the reference to ‘other non-judicial decision-making authorities’ must include a reference to the primary decision-makers in that case, the Immigration Minister and his/her delegates, and by analogy other primary decision-makers who do not adjudicate. Adjudication, therefore, cannot be a necessary element of judicial power.
The overlap theory must have some limits, if the separation of powers doctrine means anything. We have just identified one limit: non-judicial decision-makers cannot be empowered to authoritatively determine the limits of their own jurisdiction. Below, we explore further the question of what courts and administrators cannot do.
1 What Courts Cannot Do
The view that adjudication is inherent in judicial power makes some sense when considering the functions that can be conferred on courts. Adjudication does seem to be an ‘essential attribute of the curial process’ that Parliament must maintain. However, this limitation does not come from the concept of judicial power, but from Ch III’s more specific requirement that federal courts be given jurisdiction with respect to ‘matters’. This had been assumed in earlier cases, and Gleeson CJ and McHugh J have recently explicitly held as much in Minister for Immigration and Multicultural Affairs and Indigenous Affairs v B. Their Honours recognised some exceptions to this principle, but said ‘such cases are nevertheless rare and are recognised only for historical reasons’. Thus, the general rule is that Ch III courts cannot perform non-adjudicatory functions, because such functions will not give rise to a ‘matter’ within the meaning of Ch III.
Another suggested limitation to what courts can do is far more problematic. Some cases have held that it is not judicial power to create or change rights and duties for the future rather than to declare and give effect to existing rights and duties. However, there are far too many examples of courts creating or changing rights and duties for that distinction to mark out a clear dividing line between judicial and administrative power. On the face of it, then, the function of changing or creating rights for the future lies in the overlap area, and Parliament has a choice whether to confer it on an administrative agency or on a court. The High Court recognised as much in Precision Data Holdings Ltd v Wills saying:
the fact that the object of the determination is to bring into existence ... a new set of rights and obligations is not an answer to the claim that the function is one which entails the exercise of judicial power. The Parliament can, if it chooses, legislate with respect to rights and obligations by vesting jurisdiction in courts to make orders creating those rights or imposing those liabilities.
The Court did put some limits to this, though, continuing
where a discretionary authority is conferred upon a court ... to be exercised ... by reference to an objective standard or test prescribed by the legislature and not by reference to policy considerations or other matters not specified by the legislature, it will be possible to conclude ... that the determination constitutes an exercise of judicial power. However, where ... the function of making orders creating new rights and obligations is reposed in a tribunal which is not a court and considerations of policy have an important part to play in the determination to be made by the tribunal, there is no acceptable foundation for the contention that the tribunal ... is entrusted with the exercise of judicial power.
The obvious corollary is that it is not ‘possible to conclude that the determination constitutes an exercise of judicial power’, if there is no ‘objective standard or test prescribed by the legislature’ to be applied. But as Kirby J said in Breckler, ‘courts must frequently apply vague and indeterminate criteria which involve imprecise conclusions, moral judgments, evaluative assessments and discretionary considerations that are nonetheless proper to their function as courts’. On the flip-side, it is a fundamental administrative law principle that administrators’ statutory discretions are constrained by objective standards at least implicit in the Act. For example, such discretions can only be exercised for the statutory purposes for which they have been conferred, and by reference only to the considerations that the Act makes relevant.
Ultimately, the dividing line between those discretions that can be given to courts, and those that can only be given to non-judicial bodies, is too fine to be maintained. Indeed, apart from R v Spicer; Ex parte Australian Builders’ Labourers’ Federation, the Court has never invalidated the conferral of a function on a Ch III court on the basis that the statutory criteria are insufficiently objective. It is unlikely to want to, for example, strike down the Family Court’s jurisdiction to make orders ‘in the best interests of the child’, despite the highly discretionary nature of the relevant criteria for the making of such orders. The function of changing rights for the future, even on the basis of highly discretionary criteria, is best understood as being within the area of overlap, and therefore, able to be conferred on either judicial or non-judicial bodies in Parliament’s discretion, subject, of course, to Ch III’s requirement that there be a ‘matter’. In practice, it is that requirement, far more than the separation of powers doctrine, that limits what courts can do.
2 What Non-Judicial Bodies Cannot Do
On the other hand, the separation of powers doctrine does significantly limit what non-judicial bodies can do. We have already identified one such limit: they cannot be empowered to authoritatively determine the limits of their own jurisdiction. Plaintiff S157 thus constitutionally entrenches a minimum level of scrutiny of administrative decision-making by an independent judiciary, ensuring the near universal availability of a basic accountability mechanism.
Whilst this principle was only recently affirmed unequivocally, the availability of judicial review has often been mentioned in the course of a finding that an administrative agency is not exercising judicial power. Indeed, developments in administrative law generally, and judicial review in particular, have provided a key backdrop to the High Court’s separation of powers jurisprudence. The intensity of judicial scrutiny of administrative action increased markedly during the course of the 20th century, but through developing the grounds of judicial review, rather than increasing the strictness of separation of powers requirements to reserve functions to the courts themselves. The resulting ‘constitutional settlement’ might be termed ‘flexibility + accountability’: the overlap theory gives Parliament wide latitude in establishing and empowering administrative agencies, while judicial review ensures accountability to the courts. Plaintiff S157 has now entrenched this settlement within the separation of powers doctrine itself.
More generally, legislative schemes conferring functions on non-judicial decision-makers usually will not be found to breach separation of powers requirements, provided they do not preclude recourse to the courts. Thus, in deciding whether judicial power has been conferred on a non-judicial body, the High Court usually identifies all the ways in which its determinations are subject to some form of judicial oversight, whether through judicial review, collateral attack, statutory appeal, or separate enforcement proceedings. What the Court will not allow is for non-judicial bodies to effectively supplant the courts, in the sense of providing a separate mechanism for determining how legal principles apply to the facts as found, without being subject to adequate mechanisms for curial oversight.
The problem is that the High Court has never been clear about what minimum level of recourse to the courts is required to avoid a finding that judicial power has impermissibly been conferred on an administrative agency. Implicitly, the Court seems to have drawn a distinction between administrative decision-making on the one hand, and administrative dispute resolution on the other. By ‘administrative decision-making’, we mean functions such as immigration, tax, and social security, in which administrative officials determine what an individual’s rights and liabilities under a statute should be, including merits review functions. By ‘administrative dispute resolution’, we mean mechanisms designed to resolve disputes between members of the community, such as between employers and employees over wages and conditions, or between those complaining of racial or sexual discrimination and those accused of having perpetrated it.
As we have seen, Plaintiff S157 held that the separation of powers doctrine means that administrative decision-makers cannot be given power to conclusively determine their own jurisdiction. This principle applies equally to both administrative decision-making (an issue in Plaintiff S157 itself) and to administrative dispute resolution bodies. But while that seems to be all that the separation of powers means for administrative decision-makers, it seems to require something more for administrative dispute resolution bodies. In Brandy, the Human Rights and Equal Opportunities Commission (HREOC) was subject to judicial review, and certainly did not have power to conclusively determine its own jurisdiction, but that was not enough to avoid it being found to be purporting to exercise judicial power.
The primary reason for this lay in the Court’s fastening on the concept of ‘enforcement’ as an exclusively judicial power. Non-judicial bodies can change legal rights and duties for the future, and can also determine that existing rights have been breached. However, starting with Alexander, the separation of powers has been interpreted to mean that such determinations cannot be made directly enforceable in a manner analogous to ‘the system of enforcement by a court’s own officials, bailiffs and sheriffs, acting under specific court orders authorising various curial processes of, for example, forfeiture, seizure, arrest, execution and sale’.
The combined effect of Alexander and Boilermakers was that the two functions of award-making and award-enforcement had to be split between a non-judicial industrial relations commission and a judicial industrial relations court respectively. In Joske, Barwick CJ criticised this outcome as ‘unnecessary ... for the effective working of the Australian Constitution’ and as leading to ‘excessive subtlety and technicality in the operation of the Constitution without ... any compensating benefit’. However, in the employment regulation field, it was a settlement that the major players could live with. Employers and organised labour are both well-resourced repeat players quite capable of navigating the legal system’s intricacies and opportunistically employing them to their advantage. In practical terms, it is hard to know whether this particular separation of functions benefited the players or disadvantaged them: was the increased cost of enforcement actions occurring in the court rather than the commission offset by higher quality decision-making? Probably not, but either way, both sets of players were affected pretty much equally.
The same cannot be said of Brandy’s practical effect. While subjecting award enforcement to curial determination equally affects both employers and employees, requiring HREOC determinations to be subject to curial redetermination only protects those accused of discrimination. If a respondent does not like HREOC’s determination that they have discriminated, then the separation of powers guarantees them a rehearing in a Ch III court before an enforceable order against them can be made. On the other hand, if a complainant does not like HREOC’s determination that they were not discriminated against, then they are not guaranteed a curial rehearing. The separation of powers doctrine only guarantees them judicial review for jurisdictional error. And this in a context of clear power and resource disparities between, for example, Indigenous people on the one hand, and the employers, property owners and businesses that discriminate against them on the other. Whilst it can readily be accepted that the separation of powers means that a body like HREOC cannot totally supplant Ch III courts, it is certainly not clear that it should protect these particular interests in these particular ways.
A further comparison is also instructive. Whilst a person accused of discrimination has a right to a curial redetermination before being made to pay, say, $2500 damages, a disappointed refugee applicant, like a disappointed HREOC complainant, is only guaranteed minimal judicial review. The doctrinal reason for this is that refugee determinations are not ‘enforceable’ in a manner directly analogous to court decisions. This ignores the very real way in which adverse determinations are enforced through mandatory detention and forcible repatriation. In principle, a hearing on the merits by an independent judiciary is more important in the case of a person being forcibly detained by the state and repatriated, possibly to their death, than in the case of a person being made by the state to pay a relatively small sum of money.
Not only are the stakes much higher for the refugee claimant than the person accused of discrimination, but the basic rationale of the separation of powers applies more strongly. The point of the separation of powers is that judicial tenure makes the judiciary relatively independent of the executive government. Yet merits review bodies decide controversies between individuals and the executive government of which they are a part. Their lack of tenure means that they are liable to be not reappointed (effectively sacked) by the very government whose decisions they are reviewing. This may not conclusively show that Ch III courts should resolve these controversies, but it does suggest that the minimum level of judicial scrutiny necessary here ought to be higher than in cases where a tribunal decides controversies between members of the community from whom they are quite separate and apart. There is no evidence, for example, that the States’ anti-discrimination tribunals are any less objective and impartial than Ch III courts, notwithstanding their members’ lack of judicial tenure.
The High Court’s preoccupation with enforcement as the relevant limit to what administrative dispute resolution bodies can do demonstrates the problems that arise when principles worked out largely in the context of employment regulation are uncritically applied in very different contexts. In principle, the separation of powers doctrine ought at least to be even-handed between administrative dispute resolution and administrative decision-making. In practice, the Court has been far stricter with the former than the latter, ignoring its own stated commitments to flexibility. It has created rigidities divorced from the doctrine’s underlying rationale, preventing Parliament from finding traditional curial methods of dispute resolution wanting in particular areas, and creating effective alternative systems. And, as we shall see in Part IV, these problems are further magnified when the principles are applied in the context of responsive regulation.
As we have shown, the High Court developed its separation of powers doctrine mainly in the context of ‘traditional’ administrative agencies exercising statutory powers. However, in recent years, Commonwealth governments have embraced a ‘responsive regulation’ approach, which suggests that public policy objectives may be better realised through industry self-regulation, with significant statutory intervention kept as a last resort, for when self-regulation fails to achieve public policy objectives. This shift is reflected in the Commonwealth’s greater willingness to engage in institutional experimentation, increasingly allowing private or quasi-private self-regulatory bodies to exercise functions that, traditionally, would be regarded as public administration. As we have already suggested, the ‘overlap’ theory was always intended to provide a foundation for institutional experimentation, and the two Federal and High Court cases considered here show the courts continuing to take a flexible approach. The question is, in a world where private or quasi-private bodies exercise public power, does flexibility come at the cost of a loss of accountability?
In Breckler, the High Court considered the nature of the power being exercised by the Superannuation Complaints Tribunal, a statutory body with power under the Superannuation (Resolution of Complaints) Act 1993 (Cth) (the ‘Complaints Act’) to review the merits of decisions made by the trustees of regulated superannuation funds. Unlike traditional ‘command-and-control’ type regulation, however, which is unilaterally imposed on regulated entities, this regulatory scheme only applies if the trustees have elected, under the Superannuation Industry (Supervision) Act 1993 (Cth) (the ‘Supervision Act’), to bring the fund under the umbrella of Commonwealth superannuation regulation. The incentive to do so is that the fund is then eligible for concessional tax treatment. Once the election is made, the Supervision Act imposes standards on trustees, which are deemed to form part of the trust instrument. One standard is that trustees comply with the Tribunal’s orders, directions or determinations. In Breckler, the Tribunal had exercised its powers under the Complaints Act, and made an order revising the trustees’ distribution of trust funds to the beneficiaries. The trustees challenged the Tribunal’s decision, on the basis that it represented an exercise of judicial power by a non-judicial body.
The High Court rejected the trustees’ argument. The majority judgment (Gleeson CJ, Gaudron, McHugh, Gummow, Hayne and Callinan JJ) found that the Tribunal was not exercising sovereign power, in the sense used by Griffith CJ in Huddart Parker. It was exercising private power because the regulatory standards were incorporated into the private law trust instrument, and the dispute resolution procedures were voluntary, since they required an election by the trustees to submit to them.
The majority also held that the Tribunal’s determinations are not conclusive, merely providing ‘a criterion by reference to which legal norms are imposed and remedies provided for their enforcement’. Those remedies include imposition of a fine on trustees failing to observe a determination, and an injunction requiring trustees to observe a determination, which can be sought by any person whose interests are affected by the Tribunal’s decision. These remedies would require the Federal Court to engage in ‘an independent exercise of judicial power’. The Tribunal was also subject to judicial review. The scheme thus maintained sufficient judicial oversight, although it is unclear what status the Tribunal’s determination has in enforcement proceedings: should the court review its merits, all questions of law, for jurisdictional error, or not at all?
While the majority’s reliance on the avenues for judicial oversight was conventional, its primary holding that the Tribunal exercised private rather than public power has an air of unreality about it. The Tribunal is a statutory body staffed by ‘officers of the Commonwealth’ and exercising statutory powers, and, as the majority acknowledged, trustees have no real choice but to opt into the regulatory scheme, given their potential liability for breach of trust if they fail to obtain the tax concessions. The majority may have felt comfortable in construing the Tribunal’s power as private because this was not critical to ensuring that intended beneficiaries could enforce Tribunal decisions. However, as the next case demonstrates, it is a short step from Breckler’s scenario to one where individuals are left with no effective means of judicial oversight of the exercise of power by a quasi-private body.
B Viper Communications
Viper Communications arose when the Australian Communications Authority (ACA) sought pecuniary penalties against two small Internet Service Providers (ISPs), Viper Communications Pty Ltd and Albury Local Internet Pty Ltd, for their failure to join the Telecommunications Industry Ombudsman (TIO) scheme (TIO scheme). They responded by challenging the scheme’s constitutional validity, arguing that its statutory underpinning, the Telecommunications (Consumer Protection and Service Standards) Act 1999 (Cth) (the ‘Service Standards Act’), confers judicial power on the TIO, a non-judicial body.
The TIO scheme was established in 1993, in the context of government policy to open up the telecommunications industry to competition. It covers standard and mobile telephone service providers, as well as ISPs, and has three key elements: a private company called Telecommunications Industry Ombudsman Limited (TIO Ltd) a TIO council appointed by the board of TIO Ltd and the office of the TIO itself.
Service providers must comply with service provider rules, one of which is compliance with the Service Standards Act. Under that Act, carriers and carrier service providers must join the TIO scheme (s 128(1)), which is binding on them (s 132). Under s 128(4), the TIO scheme must ensure that the TIO is able to investigate complaints by end-users (residential and small business customers) about the relevant services, and make determinations or give directions in relation to those complaints. The ACA can direct service providers to join the TIO scheme.
The TIO scheme is industry-funded, with each member of the scheme being charged complaint handling fees according to the number and cost of complaints made about them. This means that those service providers with the largest customer base (eg, Telstra and Optus) inevitably contribute the bulk of the TIO scheme’s operational costs.
The TIO Constitution states that the TIO scheme ‘is committed to the Principles of Accessibility, Independence, Fairness, Accountability, Efficiency and Effectiveness’. The emphasis is on dealing with a large volume of complaints in an expeditious and inexpensive way, with the TIO required to resolve complaints in a ‘fair, just, economical, informal and expeditious [way]’, and to approach this task with reference ‘to the law, good industry practice and what is fair and reasonable in all the circumstances’. Where a complaint cannot be resolved by conciliation, the TIO is empowered to make binding determinations, up to a value of $10,000. The TIO may also make non-binding recommendations up to a value of $50,000.
Complainants may request internal merits review, but there is only a limited right to reasons for decisions: ‘only such written reasons as to give effect to any decision or recommendation’. Complainants can elect to accept TIO decisions, in which case the matter is settled. If they choose not to, the service provider is released from the TIO’s decision, and it is then left to the complainant to seek an alternative remedy. However, complainants have no right to challenge TIO decisions.
In deciding that s 128 of the Service Standards Act does not confer judicial power on the TIO, Sackville J focussed on two key issues: the TIO’s decision-making processes, and the enforceability of its decisions. In relation to the first, Sackville J found that the language of s 128 left the TIO free to adopt ‘non-judicial’ methods for resolving disputes. These methods took a number of forms. First, the statute contemplated a scheme in which the TIO is required to investigate complaints, and this is a ‘function that is not readily compatible with the exercise of the judicial power of the Commonwealth’. Second, s 128 left the TIO ‘free to create norms to resolve a particular dispute or class of dispute’. Sackville J also highlighted the TIO Constitution’s provision that the TIO was to decide on the basis of ‘what is fair and reasonable in all the circumstances’ as well as ‘the law’. This recognises, he said, ‘implicitly if not explicitly, that the TIO will make many, if not all determinations ... otherwise than by applying legal principles to the facts as found’. None of these are particularly persuasive: inquisitorial bodies such as the Refugee Review Tribunal are clearly none the less subject to the separation of powers doctrine; the precedential value of Ch III court decisions lies directly in their creation of ‘norms to resolve a particular dispute or class of dispute’ and, as discussed above, courts commonly apply criteria such as ‘fair and reasonable’ in resolving disputes.
Doctrinally, Sackville J was on firmer ground in holding that TIO determinations are not directly enforceable in the way judicial determinations are. Indeed, in contrast to the position of complainants in Breckler, Sackville J held that complainants to the TIO do not have any right at all to enforce TIO determinations. Section 132 of the Service Standards Act provides ‘[a] carrier or carriage service provider who is a member of the Telecommunications Industry Ombudsman scheme must comply with the scheme’. This is reinforced by Part 30 of the Telecommunications Act 1997 (Cth) which provides that the ACA, Australian Competition and Consumer Commission or the Minister (but not a complainant) may apply for an injunction requiring a person to comply with the requirements of the Services Standards Act. Enforcement of a TIO determination is therefore a discretionary matter for the ACA, with no certainty that the ACA would always be successful, given the grant of an injunction is itself a discretionary matter for the court.
Sackville J also considered the possibility of enforcement based on the TIO’s corporate structure. Given that under s 140 of the Corporations Law the memorandum and articles of association for TIO Ltd constitute a contract between TIO Ltd and each member, and separately between each member and another member, it would be open to TIO Ltd or a service provider (but again not a complainant) to take action to have a TIO determination enforced. Again, an injunction or order for specific performance would not be automatically available but would depend on the discretion of the court in a particular case.
The practical effect of the decision in Viper Communications is to undermine the position of complainants, regardless of whether their complaint is successful or not. If they win a determination from the TIO in their favour, they have no guarantee that it will be enforced. As will be argued further below, if their complaint is dismissed, they have no right to seek judicial review of the TIO’s decision. The vulnerability of complainants is further exacerbated by the limited right to reasons for decisions, and the absence of any right to obtain relevant documents under FOI.
C Towards Responsive Regulation or Away from Accountability?
Decisions such as Breckler and Viper Communications give Parliament a clear judicial mandate to continue to be creative in designing institutions and allocating functions. They demonstrate that the High Court’s flexible approach to the separation of powers has created a legal environment in which a culture of responsive regulation can develop. This is consistent with the premise that the separation of powers should not unreasonably constrain Parliament from experimenting with institutional design.
Regulatory frameworks such as the TIO scheme reflect a sharp shift away from a traditional, interventionist model of public sector regulation, or a ‘command and control’ approach, to one that incorporates a central role for non-government bodies in giving effect to public policy objectives. This shift may bring a number of benefits, that include:
|•||Improved legitimacy and compliance stemming from industry participation in rule development and implementation (eg although the ground rules for the TIO scheme are set by statute, the detail and application are developed by the industry);|
|•||More effective rule design, by harnessing industry expertise and local knowledge;|
|•||The privatisation and internalisation of the costs of the regulatory regime (eg under the TIO scheme, service providers pay for the scheme, in proportion to the number of complaints about their services); and|
|•||More responsive and flexible regulatory frameworks (eg the TIO scheme allows for informal, expeditious and inexpensive dispute resolution).|
However, this flexibility in institutional design may come at the cost of a loss of accountability in circumstances where a private or quasi-private body is exercising public power. As Viper Communications demonstrates, as currently understood, the separation of powers doctrine only guarantees that one side of a dispute will have access to the accountability mechanism of judicial oversight.
In that case, Sackville J accepted that, in any enforcement proceedings, it would be open to the service providers to challenge TIO determinations on a number of grounds, including that the TIO exceeded its jurisdictional limit, failed to follow the procedures laid down by the TIO Constitution, or made some clear error of fact or law which is manifest on the face of the determination. Thus, enforcement proceedings provide service providers with a surrogate judicial review mechanism. On the other hand, complainants are not constitutionally guaranteed any recourse to the courts at all. As we have already shown, even if a complainant obtains a favourable determination from the TIO, they cannot enforce it themselves. Furthermore, current administrative law doctrine in Australia does not recognise a body like the TIO as subject to judicial review. It is not staffed by ‘officers of the Commonwealth’, and its decisions are not made ‘under an enactment’. A complainant might be able to succeed in a common law judicial review action in a State Supreme Court based on R v Panel on Take-Overs and Mergers; Ex parte Datafin, an influential English Court of Appeal decision holding that a private, unincorporated association, responsible for regulating takeovers in the jurisdiction of the City of London, could be subject to public law remedies. However, that case has never authoritatively been accepted in Australia. As with Brandy, it is unclear why the separation of powers doctrine should prefer the interests of powerful private interests like telecommunications service providers over consumers in this way.
Viper Communications is entirely consistent with the High Court’s separation of powers jurisprudence, and shows just how fragile it is; a doctrine that should be about guaranteeing access to the courts in fact fails to do so. In his dissenting opinion in Neat Domestic, a case not concerned with separation of powers but highly relevant to the issue here, Kirby J stated that:
The question of principle presented is whether, in the performance of a function provided to it by federal legislation, a private corporation is accountable according to the norms and values of public law or is cut adrift from such mechanisms of accountability and is answerable only to its shareholders and to the requirements of corporations law or like rules. Given the changes in the delivery of governmental services in recent times, performed earlier and elsewhere by ministries and public agencies, this question could scarcely be more important for the future of administrative law. It is a question upon which this court should not take a wrong turning.
Later in his judgment, Kirby J suggested:
The character of the decisions of bodies assigned important public functions is not determined conclusively by the structure of such bodies (for instance as private or statutory corporations), still less by arguments about the merits or demerits, advantages or disadvantages of privatisation or private sector management. In so far as such decisions derive their necessity or effectiveness, and the bodies making them derive their existence or particular functions, from federal legislation, they may involve the exercise of public power. In so far as they do this, under the Constitution, a minister must be accountable to the parliament in respect of such exercise. In turn, through the parliament, the minister, and the government of which he or she is part, are responsible to the electors.
Just as the doctrine of ministerial responsibility should constitutionally entrench a line of accountability to Parliament, we argue, the separation of powers doctrine should entrench a line of accountability to the courts. Yet, in its focus on enforcement, the High Court seems already to have taken a ‘wrong turning’, that fails to ensure this.
However, the majority left the broader question of principle open, stating that their answer in the particular circumstances of Neat Domestic should not be taken as an answer to the broader question of whether public law remedies may be imposed on private bodies. This question was affirmatively answered in the United Kingdom in 1987. Unless a similar answer is reached in Australia, how is effective accountability through judicial oversight to be maintained in a regulatory context in which non-government agencies make and implement public policy?
The separation of powers doctrine is concerned with the circumstances in which access to the courts will be constitutionally guaranteed. The underlying rationale of the doctrine is to provide accountability in governmental decision-making based on judicial independence. Parliament may happen to provide additional measures of accountability, such as judicial review under the ADJRA or a right to appeal on a question of law, but these measures are not required under the doctrine. Where such measures are not provided, the importance of the constitutional guarantee is laid bare. As we have shown, in the context of responsive regulation, where quasi-private or private non-statutory bodies supplant traditional administrative bodies, traditional avenues of judicial review may not be available. In such a setting, the High Court’s jurisprudence on separation of powers fails to provide equal access to a minimum level of judicial oversight for complainants and respondents. Although respondents must have some access to the courts in relation to a decision through enforcement proceedings (as in Brandy, Breckler and Viper Communications), a complainant may be locked out of judicial review. This is most transparent in Viper Communications, since the effect of the decision is that a complainant either is unable to enforce a TIO determination in their favour, or has no recourse to judicial review of an unfavourable decision.
In Plaintiff S157, the High Court established that a minimum level of judicial review of administrative decision-making is a fundamental constitutional requirement. But this decision has less purchase in a regulatory context where public officials or statutory agencies no longer exercise public power. The changed regulatory context means the separation of powers doctrine, as currently articulated by the High Court, fails to guarantee equal access to a minimum level of judicial oversight. This accountability hole could be addressed in two important ways. One would be for the High Court to accept that public law remedies may be applied to private bodies exercising powers of public administration. As well, the separation of powers doctrine could be re-shaped in a way that removes enforceability as a key indicator of the exercise of judicial power in a non-judicial context. Cases establishing the importance of enforceability in a non-judicial context were decided at least 50 years ago, in different regulatory contexts. Their application today produces results divorced from the underlying rationale of the separation of powers, and which protect competing interests unequally, as our analysis of Brandy and Viper Communications demonstrate. A separation of powers doctrine that allows enforcement of decisions by non-judicial agencies exercising public power, and that supports access to judicial review of decisions of those agencies regardless of their provenance, would be more consistent with the underlying rationale of the separation of powers doctrine.
[∗] Respectively Senior Lecturer and Associate Lecturer, Griffith Law School.
 (2003) 211 CLR 476 (‘Plaintiff S157’).
 Plaintiff S157 (2003) 211 CLR 476, 511 (Gaudron, McHugh, Gummow, Kirby and Hayne JJ). See also Duncan Kerr and George Williams, ‘Review of Executive Action and the Rule of Law Under the Australian Constitution’ (2003) 14(4) Public Law Review 219.
 Australian Constitution s 75(v); ASIC v Edensor Nominees Pty Ltd (2001) 204 CLR 559, holding that ASIC is ‘the Commonwealth’ for the purposes of the Australian Constitution s 75(iii).
 Geoff Airo-Farulla, ‘Administrative Law and Governance’ in Chris Finn (ed), Administrative Law for the New Millenium (2000) 268; Geoff Airo-Farulla, ‘Internet content, mandated self-regulation and institutionalizing values’ in Ian Holland and Jenny Fleming (eds), Government Re-formed: Values and New Political Institutions (2003) 205.
 Plaintiff S157/2002 (2003) 211 CLR 476, 511 (Gaudron, McHugh, Gummow, Kirby and Hayne JJ).
 See generally Ian Ayres and John Braithwaite, Responsive Regulation: Transcending the Deregulation Debate (1992).
 The Task Force on Industry Self-Regulation, Industry Self-Regulation in Australia: A Directory of National Schemes Designed to Improve Market Outcomes for Consumers <http://www.selfregulation.gov.au /dir_nat_schemes/directory.asp> at 5 May 2004, identifies over 40 such schemes, covering a wide range of industries including pharmaceuticals, travel, telecommunications and internet, retail, finance, direct marketing, broadcasting and media, food and housing.
 (1999) 197 CLR 83 (‘Breckler’).
 (2001) 183 ALR 735 (‘Viper Communications’).
 Noel Butlin, ‘Colonial Socialism in Australia 1860-1900’ in Hugh Aitken (ed), The State and Economic Growth (1959) 26, 27; Noel Butlin, Alan Barnard and Jonathon Pincus, Government and Capitalism: Public and Private Choice in Twentieth Century Australia (1982) 13-18.
 See Paul Finn, Law and Government in Colonial Australia (1987); Roger Wettenhall, ‘Administrative Boards in Nineteenth Century Australia’ (1963) 22 Public Administration 255; Roger Wettenhall, ‘The Ministerial Department: British Origins and Australian Adaptions’ (1975) 32 Public Administration 233.
 See, eg, the relatively few examples given in Federal Commissioner of Taxation v Munro (1926) 38 CLR 153, 179 (Isaacs J).
 Huddart, Parker & Co Pty Ltd v Moorehead (1909) 8 CLR 330 (‘Huddart Parker’).
 Butlin, Barnard and Pincus, above n 10, 151-152.
 Agriculture also came to be protected through statutory marketing schemes, an interesting experiment in institutional design that was rigorously tested against the Australian Constitution under s 92 rather than on separation of powers grounds.
 Eg, Excise Tariff Act 1921 (Cth) s 2.
 Butlin, Barnard and Pincus, above n 10, ch 4.
 Frank Castles, Australian Public Policy and Economic Vulnerability: A Comparative and Historical Perspective (1988) 137.
 Ibid ch 6; Frank Castles, ‘Welfare and Equality in Capitalist Societies: How and Why Australia was Different’ in Richard Kennedy (ed), Australian Welfare: Historical Sociology (1989) 56; Butlin, Barnard and Pincus, above n 10, ch 6.
 Glyn Davis et al, Public Policy in Australia (2nd ed, 1993) 32.
 (1918) 25 CLR 434 (‘Alexander’).
 (1931) 46 CLR 73 (‘Dignan’).
 (1925) 35 CLR 422.
 (1926) 38 CLR 153 (‘Munro’).
 See, eg, Silk Bros Pty Ltd v State Electricity Commission (Vic) (1943) 67 CLR 1; Rola Co (Aust) Pty Ltd v Commonwealth (1944) 69 CLR 185.
 Australian National Airways v Commonwealth (1945) 71 CLR 29 (‘ANA Case’); Attorney-General (Vic) v Commonwealth (1945) 71 CLR 237 (‘First Pharmaceutical Benefits Case’); Bank of New South Wales v Commonwealth (1948) 76 CLR 1 (‘Bank Nationalisation Case’); Federal Council of the British Medical Association in Australia v Commonwealth (1949) 79 CLR 201 (‘Second Pharmaceutical Benefits Case’). See generally Brian Galligan, Politics of the High Court (1987) ch 4; Geoff Airo-Farulla, ‘Labor and the Politics of Constitutional Interpretation’ in Glenn Patmore (ed), The Big Makeover: A New Australian Constitution (2001) 224.
 (1955) 94 CLR 254 (‘Boilermakers’). R v Davison (1954) 90 CLR 353 (‘Davison’) a challenge to the powers of the long-standing Bankruptcy Court, was the other notable separation of powers case in this period.
 Nationwide News Pty Ltd v Wills (1992) 177 CLR 1, 69 (Deane and Toohey JJ); Lange v Australian Broadcasting Corporation (1997) 189 CLR 520, 557.
 New South Wales v Commonwealth (1915) 20 CLR 54 (‘Wheat Case’); Boilermakers (1955) 94 CLR 254. Separation between the Commonwealth Executive and the High Court has also been seen as required by the need for an independent arbiter of the Commonwealth and State Parliaments’ respective legislative powers under the Constitution.
 Australian Constitution ss 75-77.
 Under s 76(i), which provides that Parliament may confer original jurisdiction on the High Court ‘in any matter ... arising under this Constitution, or involving its interpretation’ – including the interpretation of the phrase ‘judicial power’ in s 71.
 For the sake of clarity, we should state that, in line with the evaluative premises we identified above, we do not see this flexibility as necessarily a bad thing.
 Cf Breckler (1999) 197 CLR 83, 126 (Kirby J).
 See generally William Leuchtenburg, The Supreme Court Reborn: the Constitutional Revolution in the Age of Roosevelt (1995). For an introductory overview of this infamous US constitutional crisis, see Constitutional Rights Foundation, FDR Tries to ‘Pack’ the Supreme Court (1994) <http://www.crf-usa.org/bria/bria10_4.html> at 5 May 2004.
 Boilermakers (1955) 94 CLR 254, 276 (Dixon CJ, McTiernan, Fullagar and Kitto JJ).
 (1931) 46 CLR 73, 117.
 Ibid 91. See also Munro (1926) 38 CLR 153, 178-9 (Isaacs J); R v Federal Court of Bankruptcy; Ex parte Lowenstein (1938) 59 CLR 556, 557. For modern statements, see Breckler (1999) 197 CLR 83, 128 (Kirby J); Abebe v Commonwealth (1999) 197 CLR 510 (‘Abebe’) 581 (Kirby J).
 R v Trade Practices Tribunal; Ex parte Tasmanian Breweries Pty Ltd (1970) 123 CLR 361 (‘R v Trade Practices Tribunal’) 373 (Kitto J), quoting Labour Relations Board of Saskatchewan v John East Iron Works Ltd  AC 134, 148 (Privy Council). Further High Court judicial statements to the same effect are too numerous to catalogue.
 Munro (1926) 38 CLR 153, 177.
 (1995) 183 CLR 245 (‘Brandy’).
 Ibid 267.
 See, eg, R v Quin; Ex parte Consolidated Foods Corp (1977) 138 CLR 1.
 See, eg, Luton v Lessels (2002) 210 CLR 333.
 (1909) 8 CLR 330.
 Ibid 357-8.
 (1915) 20 CLR 54.
 (1909) 8 CLR 330, 357.
 (1915) 20 CLR 54, 63 (Griffith CJ), 70-71 (Barton J).
 Ibid 87.
 (1926) 38 CLR 153, 177.
 Shell Company of Australia Ltd v Federal Commissioner of Taxation (1930) 44 CLR 530, 541-2, 545.
 See, eg, Kirby J’s partial review of the authorities in Pasini v Mexico (2002) 209 CLR 246, 265-7.
 (1970) 123 CLR 361, 398.
 Drake v Minister for Immigration and Ethnic Affairs (1979) 46 FCR 409.
 (2003) 211 CLR 476.
 Ibid 511 (Gaudron, McHugh, Gummow, Kirby and Hayne JJ).
 Ibid 505 (citation omitted, emphasis added), 512 (Gaudron, McHugh, Gummow, Kirby and Hayne JJ).
 Polyukhovich v Commonwealth (1991) 172 CLR 501, 607 (Deane J), 703 (Gaudron J).
 See, eg, In re Judiciary Act and Navigation Act (1921) 29 CLR 257; Re Wakim; Ex parte McNally (1999) 198 CLR 511, 542 (Gleeson CJ); Abebe (1999) 197 CLR 510, 534 (Gleeson CJ and McHugh J), 574 (Gummow and Hayne JJ).
  HCA 20 (Unreported, Gleeson CJ, McHugh, Gummow, Kirby, Hayne, Callinan and Heydon JJ, 29 April 2004).
 Ibid , [18-20].
 Ibid .
 See, eg, Alexander (1918) 25 CLR 434, 463 (Isaacs and Rich JJ); R v Marks; Ex parte Australian Building Construction Employees and Builders Labourers’ Federation (1981) 147 CLR 471 (‘Marks’) 488 (Mason J).
 For example, Family Court property and spouse maintenance orders, and parenting and child maintenance orders, under Part VIII and Part VII of the Family Law Act 1975 (Cth) respectively; sequestration orders under s 43 of the Bankruptcy Act 1966 (Cth) which, amongst other things, vests the bankrupt’s property in another person (s 58); and the courts’ powers under the Child Support (Assessment) Act 1989 (Cth) upheld in Luton v Lessels (2002) 210 CLR 333.
 (1991) 173 CLR 167 (‘Precision Data’).
 Ibid 190-91.
 Ibid 191 (citation omitted, emphasis added), see also 189.
 This phrase came from R v Trade Practices Tribunal (1970) 123 CLR 361, 377-8 where Kitto J said that such determinations are ‘foreign to the nature of judicial power’. See also Brandy (1995) 183 CLR 245, 268 (Deane, Dawson, Gaudron and McHugh JJ).
 Breckler (1999) 197 CLR 83, 126 (citation omitted). For example, compare the ‘judicial’ criteria in R v Joske, Ex parte Australian Building Construction Employees and Builders Labourers Federation (1973) 130 CLR 87 (‘Joske’) with the ‘non-judicial’ criteria in R v Trade Practices Tribunal (1970) 123 CLR 361 and Precision Data (1991) 173 CLR 167.
 R v Toohey; Ex Parte Northern Land Council (1981) 151 CLR 170.
 Minister for Aboriginal Affairs v Peko-Wallsend Ltd (1986) 162 CLR 24.
 (1957) 100 CLR 277.
 Family Law Act 1975 (Cth) s 68F. It is not necessary here to examine the precise scope of this jurisdiction in terms of which children it applies to, and whether such orders can be made in respect of persons other than their parents: see Minister for Immigration and Multicultural and Indigenous Affairs v B  HCA 20 (Unreported, Gleeson CJ, McHugh, Gummow, Kirby, Hayne, Callinan and Heydon JJ, 29 April 2004).
 Space precludes a full analysis of the ‘persona designata’ and ‘incompatibility doctrine’ cases here. These cases examine whether functions given to Ch III judges in their ‘personal capacity’ (Drake v Minister for Immigration and Ethnic Affairs (1979) 46 FLR 409; Hilton v Wells (1985) 157 CLR 57; Grollo v Palmer (1995) 184 CLR 348; Wilson v Minister for Aboriginal and Torres Strait Islander Affairs (1997) 189 CLR 1) or to State Supreme Courts (Kable v Director of Public Prosecutions (NSW) (1995) 189 CLR 51) are inconsistent with their exercise of Commonwealth judicial power. They assume that the functions in question are not judicial power, and therefore could not be directly conferred on Ch III courts.
 Judicial review for jurisdictional error.
 See, eg, R v Trade Practices Tribunal (1970) 123 CLR 361, 376 (Kitto J); Marks (1981) 147 CLR 471, 489 (Mason J).
 See Geoff Airo-Farulla, ‘Rationality and Judicial Review of Administrative Action’ (2000) 24(3) Melbourne University Law Review 543.
 R v Coldham; Ex parte Australian Workers’ Union (1982) 153 CLR 415, 419 (Mason ACJ and Brennan J) 426-8 (Deane and Dawson JJ).
 Under the Administrative Decisions (Judicial Review) Act 1977 (Cth), under s 39B of the Judiciary Act 1903 (Cth) and under Australian Constitution s 75(v).
 Luton v Lessels (2002) 210 CLR 333, 389 (Callinan J).
 Joske (1973) 130 CLR 87, 90 (Barwick CJ).
 Alridge v Booth (1988) 80 ALR 1. In Brandy, as well as holding that HREOC could not make enforceable orders, the Court clearly assumed that a Ch III court would have to conduct a de novo hearing before making such an order. As pointed out above, it clearly was not sufficient for HREOC’s determination to be subject to judicial review for jurisdictional error or error of law more generally.
 The amount the plaintiff was ordered to pay in Brandy.
 We reject as empirically unsustainable any implicit assumption that the kinds of issues to be determined in administrative dispute resolution proceedings are necessarily more complex than those in administrative decision-making, so as to require some special decision-making expertise that courts can claim over administrative tribunals.
 For example, in 1997 the then Immigration Minister, Philip Ruddock, was reported to be ‘determined to overhaul the “independent” [Refugee Review Tribunal], claiming it operates inefficiently and allows too many people to gain asylum on spurious grounds. As part of his overhaul, last week Mr Ruddock dumped 16 of the 46 members’: Michael Millett, ‘Asylum Granted Despite Order’ Sydney Morning Herald (Sydney) 5 June 1997, 3. Ministerial pressure on the RRT, including through the power to not reappoint, is also reported in Melissa Fyfe, ‘Inside the Refugee Tribunal’ The Age (Melbourne) 9 November 2001, 17. Statistics linking the RRT’s set aside rate with the length of members’ appointment have been compiled by Mary Crock, ‘Of Fortress Australia and Castles in the Air: The High Court and the Judicial Review of Migration Decisions’ (2000) 24 Melbourne University Law Review 190, 215; cited in Re Refugee Review Tribunal; Ex parte Aala (2000) 204 CLR 82, 128 (Kirby J).
 (1999) 197 CLR 83.
 Superannuation Industry (Supervision) Regulations 1994, reg 13.17B.
 (1909) 8 CLR 330, 357.
 Breckler (1999) 197 CLR 83, 110-11.
 Ibid 111.
 Supervision Act s 34.
 Supervision Act s 315(3).
 (1999) 197 CLR 83, 111 quoting Brandy (1995) 183 CLR 245, 261.
 Ibid 107-8.
 Ibid 111.
 (2001) 183 ALR 735 (Sackville J).
 For a more detailed account see Viper Communications (2001) 183 ALR 735, 739-747 (Sackville J).
 Telecommunications Act 1997 (Cth) s 101(1).
 Service Standards Act 1999 s 129(2); Telecommunications Industry Ombudsman, 2002/2003 Annual Report (2003) 27 <http://www.tio.com.au/publications/annual_reports/ar2003/annual_2003download.htm> at 10 February 2004.
 Service Standards Acts 1999 s 130(1).
 Telecommunications Industry Ombudsman, Memorandum and Articles of Association, cl 4 <http://www.tio.com.au/members/library/documents/memoandarticles.pdf> at 10 February 2004.
 Telecommunications Industry Ombudsman, Constitution, cl 2A.1 <http://www.tio.com.au/
members/library/documents/constitution.pdf> at 10 February 2004.
 Ibid cl 5.1.
 Ibid cl 2A.2.
 Ibid cl 6(1)(a).
 Ibid, cl 6.2.
 Telecommunications Industry Ombudsman, TIO Policies and Procedures, 66 <http://www.tio.com.au/policies/procedures.pdf> at 10 February 2004.
 Ibid cl 6.3.
 Ibid cl 6.1.
 Ibid cl 6.1.
 Viper Communications (2001) 183 ALR 735, 754 -755.
 Ibid 755.
 Ibid 758.
 Ibid 755.
 See especially Telecommunications Act 1997 (Cth) s 564.
 And, of course, the TIO is controlled by the same industry members whose actions it may be required to review in individual cases.
 Airo-Farulla, ‘Internet content, mandated self-regulation and institutionalizing values’, above n 4, 206.
 Viper Communications (2001) 183 ALR 735, 754. This would be consistent with the High Court’s decision in Plaintiff S157 (2003) 211 CLR 476: see above n 75ff.
 Parliament could decide, as part of the statutory scheme, to provide complainants with a right of appeal to the courts. However, this is beside the point as our focus here is on what avenues of recourse to the courts are constitutionally guaranteed by the separation of powers doctrine.
 For the purposes of Australian Constitution s 75(v) and Judiciary Act 1903 (Cth) s 39B. Nor, if it matters, is the TIO the ‘Commonwealth’ for the purposes of s 75(iii): ASIC v Edensor Nominees Pty Ltd (2001) 204 CLR 559.
 For the purposes of the Administrative Decisions (Judicial Review) Act 1977 (Cth) (‘ADJRA’): see Neat Domestic Trading Pty Ltd v AWB Ltd (2003) 198 ALR 179 (‘Neat Domestic Trading’).
  QB 815 (‘Datafin’).
 In Neat Domestic Trading (2003) 198 ALR 179, 192-193 (McHugh, Hayne and Callinan JJ), the question whether Datafin should be followed in Australia was expressly left open, cf 206 (Kirby J).
 Neat Domestic Trading (2003) 198 ALR 179, 195-196 (citations omitted).
 Ibid 202 (citations omitted).